Hastaların elecıve kabüllerinin markov karar proseslerıyle dınamık tahsısıne yeni bir yaklaşım

Uncontrolled patient flow and variability in hospital occupancy cause treatment delays, emergency blockages, cancellations, diversions and patient elopement. Efficient bed management through controlling inpatient admissions plays a significant role in decreasing waiting time, maintaining lower system occupancy and improving service delivery. Currently, for the sake of tractability, most researchers focus on a single ward, but unfortunately, in highly utilized hospitals, transferred patients from other wards make the admission gateway undesirable or infeasible both for patients and hospital managers. In this research, to obtain a better result, we allow two hospital wards to interact with each other and transfer their patients to the other's ward while there is no free bed in their own ward. We first present an optimal admission policy using a Markov decision process model, and then show some monotonicity results for optimal policies. In addition, we develop a set of sufficient conditions, which ensure that a patient type should be admitted or not, regardless of the system congestion level. Using the Markov decision process framework on inpatient admission, we show that leveraging controls to stabilize patient flow through the hospital wards can mitigate variability in hospital occupancy. This work is based on an actual private hospital concerned with resource allocation decisions and patient satisfaction improvements.